The North Carolina State Abortion fund paid for abortions for very poor women after the Hyde Amendment stopped the use of federal Medicaid funds for abortions in 1978. From 1980 through 1992, this state fund ran out of money part way through five fiscal years. In the earlier phase of this project, the investigators estimated the effects of these five temporary interruptions in abortion funding on births, abortions, birth weight and gestational age at abortion using records of all live births, abortions and fetal deaths in North Carolina since 1980. Both abortions and births to poor women were significantly affected. They propose to continue this study by examining the effects of two subsequent significant shifts in the state funding of abortions. Both of these analyses are strengthened by the addition of comparable individual-level data from Virginia and South Carolina. These data provide control populations for the analyses. First, they propose to examine the effects on abortions and births of the two-year expansion of the state abortion fund in 1993 and 1994. The program expansion included complete funding, increased payments to providers, and liberalized eligibility requirements. There is no other case of expanding the use of state funds for poor women over the past fifteen years. Second, the investigators propose to examine the effects of permanent loss of state funding for abortions, a policy shift that followed the expansion by two years. They hypothesize not only an effect of the loss of funding, but also a substantively different effect of the permanent loss for funding relative to the episodes of temporary loss of funding. To differentiate between the effects of the temporary interruptions in abortion funding and the permanent loss of funding on pregnancies, abortions and births they first compare effects of the policy changes in North Carolina with the trends in Virginia and South Carolina. They turn then to more refined test of the hypothesis that the probabilities that pregnancies end in abortions or births are related to the duration of time without funding.